Low Self-Control, Peer Rejection, Reactive Criminal Thinking, and … · 2017-08-26 · Low...

Low Self-Control, Peer Rejection, ReactiveCriminal Thinking, and Delinquent Peer Associations:Connecting the Pieces of the Crime Puzzle

Glenn D. Walters1

Received: 22 October 2015 /Revised: 25 February 2016 /Accepted: 2 March 2016 /Published online: 12 March 2016# Springer International Publishing AG 2016

AbstractPurpose The present series of studies were designed to test the control model ofcriminal lifestyle development which integrates aspects of low self-control, generalstrain, differential association, and criminal thinking.Methods Participants for the first study were 411 boys from the Cambridge Study ofDelinquency Development, and participants for the second study were 3817 childrenfrom the National Longitudinal Survey of Youth-Child (NLSY-C) sample.Results In the first study (Cambridge), peer-rated popularity (peer rejection) andteacher-rated low self-control were cross-lagged, with results showing that while lowself-control predicted peer rejection, peer rejection did not predict low self-control. Inthe second study (NLSY-C), findings revealed that (1) peer rejection predicted deviantpeer associations but not vice versa, (2) delinquency and reactive criminal thinkingmediated the peer rejection–peer delinquency relationship, and (3) negative affect(depression, anxiety, loneliness) alone did not mediate the peer rejection–peer delin-quency relationship nor did it alter the indirect effects of delinquency and reactivecriminal thinking on this relationship.Conclusions The results of these two studies suggest that theoretical integration ispossible and that reactive criminal thinking plays an important role in mediatingrelationships involving such traditional criminological variables as low self-control,strain created by peer rejection, and peer delinquency.

Keywords Low self-control . Peer rejection . Peer delinquency . Reactive criminalthinking

J Dev Life Course Criminology (2016) 2:209–231DOI 10.1007/s40865-016-0028-3

* Glenn D. [email protected]

1 Department of Criminal Justice, Kutztown University, Kutztown, PA 19530-0730, USA

http://crossmark.crossref.org/dialog/?doi=10.1007/s40865-016-0028-3&domain=pdf

Introduction

The state of theory in criminology and criminal justice has frequently been described asfragmented and disjointed [1]. A principal reason for this is the relatively large numberof single-variable theories used to explain crime [2]. Some of these theories—differ-ential association [3] and the general theory of crime [4], in particular—have madesignificant and lasting contributions to criminological scholarship and practice. Still, nosingle construct, regardless of how powerful or important, can fully account for abehavior as complex as crime. Several scholars have called for theoretical unification incriminology and criminal justice [5, 6]. These calls have gone largely unheeded, notbecause researchers do not see the need for integration but because of problemsassociated with the unification process itself [7]. Like the pieces of a puzzle, thesesingle-variable theories require integration to realize their full potential. This meansfinding a way to integrate the different theories, identifying a unifying set of principlesor conceptual mechanisms, and then testing these principles or mechanisms. Liska et al.[8] describe three types of integration: side-by-side (horizontal), end-to-end (sequen-tial), and up-and-down (deductive). The current paper utilizes the end-to-end orsequential approach to integration whereby causal conditions are ordered along acontinuum of proximal and distal relationships. Whether some of these putative causesare properly ordered is the research question that guided the two studies presented inthis paper.

Walters [9] offers an overarching theory of criminal behavior in which statisticalprocedures (mediation and moderation, primarily) are used to integrate concepts fromdifferent criminological theories. This theory is comprised of two developmentalmodels: a moral or proactive model and a control or reactive model. According toWalters [9], the moral model captures the planned, calculated, and instrumental aspectsof crime and the control model represents the impulsive, irresponsible, and emotionalaspects. Although these two models have overlapping features, they can and have beenstudied separately. Figure 1 lays out the principal constructs and pathways found in thecontrol model of criminal lifestyle development. The biological (disinhibited or diffi-cult temperament: [10, 11]) and environmental (weak parental monitoring and poorparental socialization: [12]) antecedents of the control model, despite their absencefrom Fig. 1, are of cardinal significance in the development of low self-control [4], thecore concept in the control model. Low self-control gives rise to delinquency through aprocess of weak behavioral restraint [13, 14], and weak behavioral restraint can lead toweak cognitive restraint or what is known in lifestyle theory as reactive (impulsive,

Strain (Peer

Rejection)

Delinquency

Delinquent

Peer

Associations

Low

Self-Control

Reactive

Criminal

Thinking

Delinquency Delinquency

Delinquency

Reactive

Criminal

Thinking

Proactive

Criminal

Thinking

Delinquency

Fig. 1 Control model of criminal lifestyle development

210 G. D. Walters

emotional) criminal thinking. Delinquency gives rise to a number of constructs centralto the control model of lifestyle development, to include reactive criminal thinking,strain, and delinquent peer associations (peer selection effect). Each of these variableswas selected for inclusion in the control model because of their relationship todelinquency and each other. Overlap with the moral model is indicated by the presenceof proactive criminal thinking in the latter stages of the model, stemming, in this case,from exposure to delinquent peers (peer influence effect).

Prior research is the foundation upon which the control model of criminal lifestyledevelopment is based. A number of studies have investigated specific pathways in thecontrol model, and several have tested a major model assumption, namely, that lowself-control and reactive criminal thinking are not one in the same. Low self-control, asconceptualized by Gottfredson and Hirschi [4], is behavioral in nature and comparableto behavioral impulsivity [15]. Reactive criminal thinking, as conceptualized byWalters [9], is cognitive in nature and comparable to criminal attitudes and beliefs[16]. Ericson and Carriere [2] insinuate that one reason for the surfeit of single-variabletheories in criminology is a tendency on the part of some theorists to be overly broad intheir theorizing, resulting in what might be called compound constructs containingseveral related constructs. There is also a tendency on the part of traditional crimino-logical theory to ignore or at least downplay developmental and situational influenceson behavior, a trend that has only recently been reversed with the introduction ofdevelopmental and life-course theories of crime [17]. One could argue that Gottfredsonand Hirschi’s [4] conceptualization of low self-control is one example of overly broadtheorizing that ignores developmental and situational (with the exception of opportu-nity) factors. The control model of criminal lifestyle development seeks to reverse thistendency by restricting its definition of self-control and its associated situational anddevelopmental influences to behavioral patterns (i.e., behavioral impulsivity), whileconceptualizing the cognitive aspects as an entirely new construct (i.e., reactive crim-inal thinking).

Lifestyle theory holds that low self-control precedes reactive criminal thinking, andresearch denotes that while low self-control predicts reactive criminal thinking, reactivecriminal thinking does not predict low self-control [10]. Walters [10] also discoveredthat reactive criminal thinking mediated the relationship between low self-control anddelinquency, but low self-control did not mediate the relationship between reactivecriminal thinking and delinquency. These results suggest that low self-control is theoriginal problem, with roots in behavioral disinhibition, and that reactive criminalthinking arises from low self-control and delinquency (see also, [18]) and then servesas a mediator of relationships between initial low self-control or behavioral impulsivityand subsequent delinquency and crime. Cognitive variables like reactive criminalthinking often makes excellent mediators of behavior and social conditions [19].Furthermore, research indicates that reactive criminal thinking is capable of mediatingboth the peer selection effect (participant delinquency leading to delinquent peers: [20])and crime continuity (past delinquency leading to future delinquency: [21]). Thepresence of a reciprocal relationship between delinquency and reactive criminal think-ing has received empirical support [21] and is consistent with Thornberry’s [6] reci-procity model of delinquency.

The conceptual health and relevance of the control model depend on furthertheoretical integration and the inclusion of additional evidence-based criminological

Peer Rejection and Peer Associations 211

constructs. One popular and well-validated criminological theory that may fit nicelyinto the framework of the control model of criminal lifestyle development is generalstrain theory. According to the author of general strain theory [22], certain losses,deprivations, and frustrations create strain which, in turn, lead to criminal adaptations asthe individual tries to correct the situation, ease the strain, exact revenge, protectpositive stimuli, or neutralize negative stimuli. The key, however, is the emotionalresponse. Research indicates that strain creates a number of emotional reactions, onlysome of which lead to crime. Several studies, in fact, have shown that while anger oftenleads to criminal offending, other forms of negative affect, such as depression, anxiety,and loneliness, have little to no effect on subsequent offending behavior [23–25].Although subjective or cognitive appraisals of strain are more reliably associated withspecific emotional responses than objective strain [26], there are situations in whichstrain and its associated emotional concomitants are likely to occur in the presence oflow self-control [27]. Peer rejection may be one such situation [28].

Studies show that children with low self-control are more apt to be rejected by theirpeers than children with high self-control [29–31], that rejected children are more proneto delinquent behavior than non-rejected children [32–35], and that peer rejection maylead to delinquent peer associations [36, 37]. Only two of these studies, however,employed a mediational design. Vitaro et al. [37], for one, determined that peerrejection and peer disruptiveness mediated the early disruptive behavior (low self-control)–violent delinquency relationship but only when peer rejection preceded peerdisruptiveness. In a study using the same sample and some of the same variables as thecurrent study 2, Chapple [30] determined that peer rejection and peer delinquency eachmediated the low self-control–delinquency relationship. Both mediational analyseswere limited by the fact that they evaluated indirect (mediated) effects using either acomparative fit index or normal theory approach. Because the distribution of indirecteffects is nearly always non-normal, these methodologies frequently produce mislead-ing results [38]. A better approach would be to evaluate the significance of indirecteffects using bias-corrected bootstrapped confidence intervals [39].

Piecing the research on peer rejection, low self-control, and peer deviance intoWalters’ [9] control model of criminal lifestyle development has given birth to thecontrol model of criminal lifestyle development. As previously noted, about half of therelationships depicted in Fig. 1 have received preliminary support ([10, 20, 21]),whereas the other half, such as those involving the middle portion of the model (fromlow self-control to strain [peer rejection] to peer delinquency), have yet to be formallyevaluated. That was the purpose of the current series of studies. The first studyconsisted of a cross-lagged analysis of the low self-control–peer rejection relationshippreviously observed in Chapple’s [30] single lag study (low self-control leading to peerrejection). The alternate lag (peer rejection leading to low self-control) was used to testthe direction of the relationship between low self-control and peer rejection. A secondstudy was conducted to evaluate the next section of the control model (peer rejectionleading to peer delinquency) using a completely different set of measures and partic-ipants. In this second study, the cross-lagged relationship between peer rejection andpeer delinquency was tested, similar to what had been done previously by Vitaro et al.[37]), in an effort to evaluate the direction of the relationship and assess the indirectpathways (via delinquency and reactive criminal thinking) from peer rejection to peerdelinquency. To evaluate the specificity of the mediating effect, a non-anger measure of

212 G. D. Walters

negative affect (depression, anxiety, and loneliness) was added to the peer rejection–peer delinquency regression, with the expectation that it would correlate with peerrejection but not with peer delinquency.

Study 1

Hypothesis

The hypothesis for study 1 held that a cross-lagged analysis would reveal that while lowself-control would predict peer rejection, peer rejection would not predict low self-control.

Method

Participants All 411 boys from the Cambridge Study in Delinquent Development [40]served as participants in this first study. Most of these boys were born in 1953 and camefrom intact working class homes in South London. Sampling consisted of a census of8–9-year-old boys from seven primary schools located within a mile of the researchoffice. Data were collected in interviews conducted with the boys, their parents, theirteachers, and their peers. Official arrest/conviction records were also reviewed. Themajority of boys in this study were white in appearance and of British origin (87 %).The remaining 13 % of participants were either Afro-Caribbean or of other European(German, Irish, Swedish, Australian, French, Spanish, or Portuguese) descent.

Measures The two independent/dependent variables utilized in this study were cross-lagged over a 2-year period between the ages of 8–9 and 10–11 years. Low self-controlwas based on an evaluation of the child’s behavior by his teacher who was asked torespond to the following three dichotomous questions (yes = 2, no = 1): (1) Does helack concentration or is he restless in a way that seriously hinders his learning? (2) Doeshe have difficulties (dull, timid, dirty, untidy, attention seeking, mischievous, sexual,aggressive, troublemaker) with other children in his class? (3) Is he difficult todiscipline? Responses to these three items were summed to produce a score that couldrange from 3 to 6. The items displayed good internal consistency at both age 8–9 (meaninter-item r=0.26) and age 10–11 (mean inter-item r=0.35).

The other independent/dependent variable in this cross-lagged analysis was peerrejection. Peer rejection was measured using a sociometric evaluation of a participant’spopularity as rated by his peers. The sociometric ratings were organized into fourlevels, and ratings were assigned so that higher scores reflected lower levels of ratedpopularity: 1 = popular, 2 = average-popular, 3 = average-unpopular, and 4 = unpop-ular. Lower levels of popularity were assumed to reflect greater peer rejection.

Five control variables with potential relevance to low self-control and/or peerrejection were included in the analysis: age, race/ethnicity, family income, parentalrules, and number of friends. Age was divided into four groups: 1 = 123 to 130 monthsof age, 2 = 131 to 137 months of age, 3 = 138 to 141 months of age, and 4 = 142 to158 months of age. Race/ethnicity was coded as 1 (British) or 2 (non-British). Familyincome was assessed as comfortable (1), adequate (2), or inadequate (3). Parental ruleswere classified as average (1) or slack/rigid (2) and number of friends were assessed as


many (1), an average number (2), or few to none (3). Each control variable wascollected when the child was 8 to 9 years of age, with the first four being rated by aprofessional interviewer and the number of friends being a self-report from the child.

Procedure The design for this study involved crossing teacher-assessed low self-control and peer-assessed popularity (peer rejection) at 8–9 years of age with peerrejection and low self-control at 10–11 years of age. The two evaluations wereaccordingly separated by 2 years. Measures taken when the child was 8–9 years ofage were identified by a 1 (i.e., low self-control-1, peer rejection-1), whereas measurestaken when the child was 10–11 years of age were identified by a 2 (i.e., low self-control-2, peer rejection-2). Controlling for age, race, income, parental rules, andnumber of friends, a cross-lagged regression was calculated using the structuredequation modeling program, MPlus 5.2 [41].

Missing Data Nearly three quarters of the sample had complete data on all nine variables(n=304, 74.0 %). Of the remaining participants, 58 (14.1 %) had missing data on onevariable, 36 (8.8%) hadmissing data on two variables, and 13 (3.1%) hadmissing data onthree to five variables. Missing data were handled with full information maximumlikelihood (FIML). FIML works by estimating model parameters and standard errors forthe entire sample from known relationships between non-missing data. Studies show thatthe estimates produced with FIML are significantly less biased than those generated bymore traditional procedures like simple imputation and listwise deletion [42, 43].

Results

Means, standard deviations, and correlations for the nine variables included in study 1can be found in Table 1. Multicollinearity was not a problem in this study as evidencedby the results of a collinearity diagnostic analysis (tolerance=0.892–0.987; varianceinflation factor [VIF]=1.013–1.121). Consistent with predictions, low self-control-1leading to peer rejection-2 path was significant, whereas the peer rejection 1 leading tolow self-control-2 path was non-significant in both the Bonferroni-corrected zero-ordercorrelational (see Table 1) and structured equation modeling (see Table 2 and Fig. 2)analyses. The Raghunathan et al. [44] test for comparing dependent but non-overlapping correlations revealed a significance difference between the two zero-order correlations, Z=2.03, p<0.05.

Discussion

The current results are consistent with the hypothesis that low self-control precedespeer rejection, providing preliminary support for the low self-control/general straintheory portion of the model presented in Fig. 1. In a cross-lagged analysis of theproposed low self-control–peer rejection relationship, low self-control predicted peerrejection but peer rejection did not predict low self-control. Unfortunately, becausesociometric peer data were unavailable after age 10/11, a comprehensive measure ofdelinquency was unavailable at age 12/13, and no proxy measures of reactive criminalthinking were available for the Cambridge study, it was not possible to investigate the

214 G. D. Walters

indirect (mediated) effects proposed by the control model of criminal lifestyle devel-opment. In addition, self-control was assessed with just three items, a low score on asociometric measure of popularity does not necessarily mean the child felt rejected, andpopularity is not a particularly strong operationalization of Agnew’s [45] strain con-struct. A second study was consequently conducted on the next leg of the model(general strain leading to delinquent peer associations) using a different sample ofchildren, a behavioral measure of rejection, and three waves of data, which permittedcalculation of indirect effects.

Study 2

Hypotheses

Three hypotheses were tested in this second study:

1. Peer rejection would predict delinquent peer associations but not vice versa whencontrolling for the mediating effects of delinquency and reactive criminal thinking.

2. The direct and indirect (via delinquency and reactive criminal thinking) effects ofpeer rejection on delinquent peer associations should be significant.

3. Introduction of a non-anger negative affect (depression, anxiety, loneliness) medi-ator into the mediation analysis should not alter the significant direct and indirect(via delinquency and reactive criminal thinking) effects of peer rejection ondelinquency and should not achieve significance itself (despite a significant a pathbetween peer rejection and non-anger negative affect).

Table 1 Demographic characteristics and correlations for the nine variables in study 1

Variable n M SD Range 2 3 4 5 6 7 8 9

1. Age 411 2.46 1.10 1–4 0.04 0.06 0.02 −0.00 −0.04 0.06 −0.03 0.04

2. Race 411 1.13 0.34 1–2 0.06 0.14 0.08 0.04 −0.08 0.06 0.06

3. Family income 411 1.92 0.72 1–3 0.27* 0.06 0.16 0.14 0.14 0.16

4. Parental rules 375 1.37 0.48 1–2 0.12 0.15 0.15 0.10 0.18*

5. Number of friends 390 1.67 0.68 1–3 0.07 −0.01 0.04 0.13

6. Low self-control-1 402 3.57 0.80 3–6 0.33* 0.22* 0.28*

7. Low self-control-2 388 3.71 0.93 3–6 0.16 0.24*

8. Peer rejection-1 385 2.55 1.08 1–4 0.38*

9. Peer rejection-2 353 2.46 1.09 1–4

Variable variables included in study 1, n number of participants with non-missing data, M mean, SD standarddeviation, Range range of scores in the current sample, Age chronological age (1 = 123–130 months, 2 = 131–137 months, 3 = 138–141 months, 4 = 142–158 months), Race 1 white/British, Race 2 non-white/non-British,Family income 1 comfortable, Family income 2 adequate, Family income 3 inadequate, Parental rule 1average, Parental rule 2 slack or rigid, Number of friends 1 many, Number of friends 2 average number,Number of friends 3 few or none, Low self-control-1 low self-control at age 8/9, Low self-control-2 low self-control at age 10/11, Peer rejection-1 peer rejection at age 8/9, Peer rejection-2 peer rejection at age 10/11

*p < 0.0014 (Bonferroni-corrected alpha level; 0.05/36 correlations)


Table 2 Results of a cross-lagged regression of low self-control and peer rejection: study 1

Predictor b (95 % CI) β t p

Peer rejection-1 (outcome measure)

Age −0.035 (−0.132, 0.062) −0.035 −0.70 .482

Race 0.141 (−0.181, 0.462) 0.044 0.86 .392

Family income 0.180 (0.025, 0.334) 0.120 2.28 .022

Parental rules 0.097 (−0.151, 0.345) 0.043 0.76 .446

Number of friends 0.030 (−0.135, 0.195) 0.019 0.36 .719

Low self-control-1 (outcome measure)

Age −0.035 (−0.105, 0.034) −0.049 −0.99 .320

Race 0.037 (−0.193, 0.266) 0.015 0.31 .755

Family income 0.138 (0.027, 0.248) 0.125 2.45 .014

Parental rules 0.189 (0.015, 0.363) 0.114 2.13 .033



Age 0.045 (−0.049, 0.138) 0.045 0.93 .350

Race 0.012 (−0.290, 0.314) 0.004 0.08 .939

Family income 0.111 (−0.034, 0.257) 0.074 1.50 .134

Parental rules 0.215 (−0.019, 0.450) 0.095 1.80 .064


Low self-control-1 0.245 (0.113, 0.378) 0.180 3.64 <.001

Peer rejection-1 0.320 (0.224, 0.416) 0.317 6.51 <.001

Low self-control-2 (outcome measure)

Age 0.060 (−0.018, 0.138) 0.071 1.51 .132

Race −0.320 (−0.577, −0.062) −0.116 −2.44 .015

Family income 0.078 (−0.046, 0.202) 0.060 1.23 .220

Parental rules 0.221 (0.024, 0.417) 0.114 2.20 .028

Number of friends −0.058 (−0.192, 0.076) −0.042 −0.85 .398

Peer rejection-1 0.076 (−0.007, 0.160) 0.088 1.79 .074

Low self-control-1 0.350 (0.236, 0.464) 0.299 6.03 <.001

Peer reject-1 with LSC-1 0.165 (0.081, 0.250) 0.198 3.84 <.001

Peer reject-2 with LSC-2 0.092 (0.03, 0.181) 0.111 2.04 .042

N = 411

Peer rejection-1 (outcome measure) regression equation with peer rejection at age 8/9 as the predictedvariable, Low self-control-1 (outcome measure) regression equation with low self-control at age 8/9 as thepredicted variable, Peer rejection-2 (outcome measure) regression equation with peer rejection at age 10/11 asthe predicted variable, Low self-control-2 (outcome measure) regression equation with low self-control at age10/11 as the predicted variable, Age chronological age (1 = 123–130 months, 2 = 131–137 months, 3 = 138–141 months, 4 = 142–158 months), Race 1 white/British, Race 2 non-white/non-British, Family income 1comfortable, Family income 2 adequate, Family income 3 inadequate, Parental rules 1 average, Parental rules2 slack or rigid, Number of friends 1 many, Number of friends 2 average number, Number of friends 3 few ornone, LSC-1 low self-control at age 8/9, LSC-2 low self-control at age 10/11, Peer rejection-1 peer rejection atage 8/9, Peer rejection-2 peer rejection at age 10/11, Peer reject-1 with LSC-1 correlation between peerrejection at age 8/9 and low self-control at age 8/9, Peer reject-2 with LSC-2 correlation between peer rejectionat age 10/11 and low self-control at age 10/11, b (95 % CI) unstandardized coefficient and the lower and upperlimits of the 95 % confidence interval for the unstandardized coefficient (in parentheses), β standardizedcoefficient, t asymptotic t test, p significance level of the asymptotic t test

216 G. D. Walters

Method

Participants Data for this second study came from the National Longitudinal Surveyof Youth-Child Data (NLSY-C: [46]), a convenience sample of boys and girls born tomothers from the 1979 National Longitudinal Survey of Youth (NLSY-79). Interviewsfor the NLSY-C were conducted biennially between 1986 and 2012, although severalof the instruments used in the current study were first administered in 1992 and severalothers were not administered past age 14. Three consecutive waves of data, with 2 yearsbetween each wave, were collected starting as early as age 4 and ending as late as age14. Of the 11,512 members of the NLSY-C sample, about 40 % (N=4534) hadcomplete data on at least 5 of the 10 variables used in this study. These cases wereincluded in a preliminary analysis designed to evaluate moderation effects for age, race(white, non-white), and sex.

A review of the interaction terms in a preliminary analysis of the 10 study variablesrevealed that age moderated the peer rejection–delinquent peer and delinquency–delinquent peer relationships. The results indicated that the younger age groups (4 to8 at wave 1 and 8 to 12 at wave 3) may have been too young to demonstrate some ofthe behaviors and relationships proposed in this study. Conversely, there were nosignificant moderator effects for either race or sex. Accordingly, the sample wasrestricted to those children who were 9 or 10 years of age at wave 1 and 13 or 14 yearsof age at wave 3. Over 80 % of the eligible children from the NLSY-C were either 9 or10 years of age at wave 1 and were included in the present study (N=3817). Therewere roughly equal numbers of boys (n=1901, 49.8 %) and girls (n=1916, 50.2 %) inthe sample, and the ethnic distribution was 51.9 % white, 27.8 % black, and 20.3 %Hispanic.

Measures The independent variable for the current study was a three-item indicator ofpeer rejection measured at wave 1. Peer rejection was also measured at wave 3 for thecross-lagged analyses using these same three variables. Three items from the BehaviorProblems Index (BPI: [47]), a behavioral rating scale completed by the child’s mother,were used to create the peer rejection scale: (1) “feels/complains no one loves him/her,”(2) “has trouble getting along with other children,” and (3) “is not liked by other

Age 8-9 Years Ag e 10-11 Years

.30**

Peer Reject

Low

Self-Control

.32**Peer Reject

Low

Self-Control

Fig. 2 Cross-lagged analysis of low self-control and peer rejection in study 1. Note. Standardized betacoefficients are reported; control variables are not shown. *p < 0.05, **p < 0.001


children.” Each item was rated on a three-point scale (3 = often true, 2 = sometimestrue, 1 = not true). Scores on the peer rejection scale ranged from 3 to 9, and the scaledisplayed good internal consistency at waves 1 (mean inter-item r=0.32) and 3 (meaninter-item r=0.35).

The dependent variable for this study was a single item from the BPI assessed atwave 3 which served as a proxy for delinquent peer associations. Delinquent peerassociations were also measured at wave 1 for the cross-lagged analysis using this sameitem. The item, “hangs around with kids who get into trouble,” was rated by the child’smother using a three-point scale (3 = often true, 2 = sometimes true, 1 = not true). Thisitem, which can range in score from 1 to 3, was used previously by Chapple [30] toassess delinquent peer associations.

There were three-mediator variables included in study 2, all obtained through self-report during the wave 2 interview: delinquency, reactive criminal thinking, anddepression. Delinquency was measured with self-reported involvement in five delin-quent acts rated on a four-point scale (0 = never, 1 = once, 2 = twice, 3 = more thantwice). Children were asked: (1) How often in the last year did you hurt someone badenough that they needed a doctor? (2) How often in the last year did you takesomething from a store without paying for it? (3) How often in the last year did youdamage school property on purpose? (4) How often in the last year did you get drunk?(5) How often in the last year did you skip school without permission? Scores on thedelinquency measure ranged from 0 to 15.

The Reactive Criminal Thinking (RCT) scale consisted of six items from the NLSY-C Child Self-Administered Supplement. The child rated each statement on a four-pointLikert-type scale (4 = strongly agree, 3 = agree, 2 = disagree, 1 = strongly disagree).Each statement, along with the reactive criminal thinking style it is believed torepresent, reads as follows: (1) I often get into a jam because I do things withoutthinking (cutoff); (2) I think that planning takes the fun out of things (discontinuity); (3)I have to use a lot of self-control to keep out of trouble (cutoff); (4) I enjoy taking risks(cognitive indolence); (5) I enjoy new and exciting experiences, even if they are a littlefrightening or unusual (cognitive indolence); (6) Life with no danger in it would be toodull for me (cognitive indolence). This scale’s internal consistency was found to beadequate (mean inter-item r=0.25).

In an earlier study, Walters and Yurvati [48] determined that the six items used in thecurrent study as proxies for reactive criminal thinking correlated moderately well(r=0.43, p<0.001) with the reactive criminal thinking score from the PsychologicalInventory of Criminal Thinking Styles (PICTS: [49]) in a separate sample of partici-pants. Although none of the six items on the RCT scale used in the current studydirectly referenced crime, it should be noted that reactive criminal thinking is a measureof criminal thought process (how an offender thinks) rather than of criminal thoughtcontent (what an offender thinks). As such, direct reference to crime is unnecessary aslong as the item embodies the specific criminal thought process purportedly beingassessed by the item. In fact, Walters et al. [50] found that PICTS items making noreference to crime loaded just as well onto a latent reactive criminal thinking factor asPICTS items directly referencing crime.

The depression or non-anger negative affect scale consisted of six items from theNLSY-C Child Self-Administered Supplement that had been grouped together in theNLSY-C as a broad measure of child depression and anxiety. In completing this

218 G. D. Walters

measure, the child was asked to rate how often he or she experienced six differentemotional symptoms associated with depression (feel sad/blue, feel nervous/tense/onedge, feel happy [reverse coded], feel bored, feel lonely, and feel tired/worn out) on athree-point scale (3 = often, 2 = sometimes, 1 = hardly ever). Combining the ratingsyielded a depression index with a total score that ranged from 6 to 18. The internalconsistency of this scale was modest but acceptable for a broadband indicator ofnegative affect (mean inter-item r=0.19).

Control variables for this study included age, race (white = 1, non-white = 2), sex(male = 1, female = 2), and wave 1 peer delinquency. Cole and Maxwell [51] contendthat precursor measures of the mediator and dependent variables should be included ina longitudinal path analysis to rule out the possibility that pre-existing differences onthese measures account for the results of a mediation analysis. This was possible for thedependent variable (peer delinquency) but presented a problem for the three-mediatorvariables (delinquency, RCT, and depression). Due to the restricted time frame in whichthe Child Self-Administered Supplement was administered, data were missing forslightly more than 75 % of the wave 1 measures of delinquency, RCT, and depression.As such, analyses using the mediator precursors were treated as ancillary and reportedin a footnote.

Procedure The current study used a three-wave longitudinal design with no overlapbetween waves. It therefore qualifies as prospective in nature. Data were analyzed in afour-regression path analysis computed with MPlus 5.2. Because the dependent vari-able (wave 3 peer delinquency) was categorical, the path analysis was performed with arobust weighted least squares (WLSMV) estimator. Bias-corrected bootstrapped 95 %confidence intervals (b=5000) were used to evaluate the significance of mediatingeffects (significance indicated by a confidence interval that did not include zero) basedon research showing that this procedure is superior to the standard z test procedure inmodeling the non-normality of indirect effects [39, 52, 53]. Kenny’s [54] “failsafe ef”procedure—calculated as (rmy.x) × (sdm.x) × (sdy.x)/(sdm) × (sdy)—was used to test thesensitivity of significant mediated relationships to the obfuscating effects of unobservedcovariate confounders.

The wave 2 delinquency and wave 2 RCT variables were the mediators ofinterest in this study and formed the target pathways. The wave 2 depressionvariable served as a contrast mediator in the sense that it was expected to correlatewith wave 1 peer rejection (a path of the comparison pathway) but not with wave 3delinquent peers (b path of the comparison pathway). In the cross-lagged analysis,peer rejection and peer delinquency were crossed at waves 1 and 3 with mediationby delinquency and reactive criminal thinking at wave 2. In the two-mediatoranalysis, wave 1 peer rejection served as the independent variable, wave 3 peerdelinquency served as the dependent variable, and wave 2 delinquency and wave 2RCT served as mediator variables. In the three-mediator analysis, wave 1 peerrejection served as the independent variable, wave 3 peer delinquency served asthe dependent variable, and wave 2 delinquency, wave 2 RCT, and wave 2 depres-sion served as mediator variables.

Missing Data Complete data on all 10 variables included in this study were availablefor 1941 participants (50.9 % of the total sample). Another 1343 (35.2 %) participants


were missing data on one variable, 202 (5.3 %) participants were missing data on twovariables, and 331 participants (8.7 %) were missing data on three or four variables.Wave 3 peer rejection was the variable with the most missing data (26.4 %), followedby wave 2 depression (20.3 %), wave 2 delinquency (10.3 %), wave 3 delinquent peers(8.7 %), wave 2 RCT (4.6 %), wave 1 delinquent peers (1.1 %), and wave 1 peerrejection (0.9 %). There were no missing data for age, race, or sex. As in study 1,missing data were handled with FIML.

Results

Descriptive statistics for the 10 variables included in the current study (two independentvariables, two target mediator variables, one comparison mediator variable, two de-pendent variables, and three control variables) are listed in Table 3. As the results fromthis table indicate, nearly two thirds of the zero-order correlations were significantdespite severe restrictions in the range of the age variable and the use of a Bonferroni-corrected alpha level. Collinearity diagnostics failed to show evidence ofmulticollinearity between the eight predictor variables (tolerance = 0.884–0.990;VIF=1.010–1.131).

Cross-lagged analysis was used to assess the direction of the peer rejection–peer delinquency relationship. Comparing the two direct effects made it possibleto test the direction of the relationship accounting for age, race, sex, and thetwo intervening variables (delinquency and reactive criminal thinking at wave2). This way, the direct effect of peer rejection on peer delinquency and of peerdelinquency on peer rejection could be evaluated independent of the mechanismrepresented by the indirect effects [39]. Whereas the direct path of the targetsequence (wave 1 peer rejection leading to wave 3 peer delinquency) achievedsignificance (estimate = 0.156, 95 % BCBCI = 0.117–0.199), the direct path ofthe alternate sequence (wave 1 peer delinquency leading to wave 3 peerrejection) did not (estimate = 0.152, 95 % BCBCI =−0.100 to 0.403). Convertingthe four cross-lagged variables to a common scale (z scores) permitted use ofPreacher and Hayes’ [55] contrast test to evaluate the difference between thetwo direct effects, the results of which indicated that the target sequence wassignificantly stronger than the alternate sequence (estimate = 0.066, 95 %BCBCI = 0.003–0.127).

The results for the two-mediator path analysis of the peer rejection–peer delin-quency relationship (without the cross-lags) can be found in Tables 4 and 5. Thesefindings indicate that both wave 2 delinquency and wave 2 RCT successfullymediated the peer rejection–peer delinquency relationship (see Fig. 3). Sensitivitytesting revealed that an unobserved covariate confounder would need to correlate atleast 0.14 with both the mediator and outcome (controlling for the mediator andindependent variable) to completely eliminate the indirect effect of wave 2 delin-quency on the peer rejection–peer delinquency relationship and that an unobservedcovariate confounder would need to correlate 0.08 with both the mediator andoutcome (controlling for the mediator and independent variable) to totally eliminatethe indirect effect of wave 2 RCT on the peer rejection–peer delinquency relation-ship. An ancillary analysis controlling for pre-existing levels of delinquency and

220 G. D. Walters

Tab

le3

Dem

ographicCharacteristicsandCorrelatio

nsforthe10

Variables

instudy2

Variable

nM

SDRange

23

45

67

89

10

1.Age

3817

9.53

0.50

9–10

−0.01

0.01

−0.02

−0.04

0.04

0.05

−0.01

−0.01

0.01

2.Race

3817

1.48

0.50

1–2

0.00

0.05

0.07*

0.12*

−0.03

0.06*

0.07*

0.07*

3.Sex

3817

1.50

0.50

1–2

−0.04

−0.17*

−0.14*

−0.12*

−0.00

−0.00

−0.05

4.Peer

rejection-1

3783

3.70

1.02

3–9

0.28*

0.09*

0.08*

0.15*

0.44*

0.20*

5.Peer

delinquency-1

3774

1.13

0.36

1–3

0.11*

0.08*

0.08*

0.19*

0.28*

6.Delinquency-2

3424

0.71

1.65

0–15

0.21*

0.18*

0.12*

0.16*

7.RCT-2

3642

14.60

3.40

6–24

0.21*

0.04

0.09*

8.Depression-2

3042

10.36

2.22

6–18

0.15*

0.09*

9.Peer

rejection-3

2808

3.60

1.00

3–9

0.36*

10.P

eerdelin

quency-3

3485

1.21

0.46

1–3

Variablevariablesinthetwo-

andthree-mediatormodels,nnumberof

participantswith

non-missing

data,M

mean,SD

standard

deviation,Range

rangeof

scores

inthecurrentsam

ple,

Age

chronologicalage

inyears,Race1(w

hite),Race2non-white,Sex

1(m

ale),Sex

2(fem

ale),P

eerrejection-1maternalratingof

difficultand

rejectingpeerrelationships

atwave1,

Peerdelinquency-1

peerdelin

quentassociatio

nsatwave1,Delinquency-2

self-reportedinvolvem

entindelinquency

during

wave2,RCT-2reactivecrim

inalthinking

scoreatwave2,

Depression-2self-reporteddepression

atwave2,Peerrejection-3maternalratingof

difficultand

rejectingpeerrelatio

nships

atwave3,Peerdelin

quency-3

peerdelin

quentassociations

atwave3

*p<0.0011

(Bonferroni-correctedalphalevel,0.05/45correlations)


reactive criminal thinking (wave 1 delinquency and wave 1 RCT) achieved similarresults.1

A comparison pathway (wave 1 peer rejection leading to wave 2 depression 2leading to wave 3 peer delinquency) with a non-anger negative affect mediator (wave2 depression) was added to the path analysis to create a three-mediator model. Theresults of this analysis revealed that the comparison pathway failed to achieve signif-icance but that the indirect effects of wave 2 delinquency and wave 2 RCT on the peerrejection–peer delinquency relationship remained significant even with introduction ofa non-anger negative affect mediator (see Table 6). As predicted, the a path from peerrejection to depression was significant, the b path from depression to peer delinquencywas not significant (see non-bolded coefficients in Fig. 3), and the overall indirect effectwas non-significant. What this means is that peer rejection generates high levels ofnegative affect in the form of anxiety, depression, and loneliness but that these emotionsdo not necessarily lead to delinquent peer associations.

Discussion

The results of this second study provide preliminary support for the strain–delinquentpeer association portion of the control model of criminal lifestyle development depictedin Fig. 1. This section of the control model—which had not been formally tested priorto the current investigation—was assessed against three hypotheses, all of whichreceived support in this study: (1) the directional effect appeared to go from peerrejection (strain) to peer delinquency rather than from peer delinquency to peerrejection; (2) delinquency and reactive criminal thinking mediated the peer rejection–peer delinquency relationship; and (3) including non-anger negative affect (depression,anxiety, and loneliness) as a mediator in a three-mediator model did not alter thesignificant direct and indirect effects observed in the two-mediator model. This indi-cates that strain, as measured in the current study by peer rejection, exerts both a directand indirect effect on delinquent peer associations although the mechanism of effect issomething other than depression, anxiety, or loneliness. Other features of negativeaffect (anger and hostility within the context of reactive criminal thinking) and delin-quent behavior (in the formation of a peer selection effect) along with variables not yetidentified appear to be what links peer rejection to delinquent peer associations.

One could argue that the simple mediation effects of delinquency and particularlyreactive criminal thinking on the peer rejection–peer delinquency relationship weresmall and therefore unimportant. There are several factors that should be considered,

1 In an attempt to control for pre-existing differences in delinquency and reactive criminal thinking, precursormeasures for the mediators (wave 1 delinquency and wave 1 RCT) were added to the two-mediator model.The WLSMVanalysis would not run, however, due to the fact that the weight matrix portion of the categoricalwave 3 peer delinquency variable was non-invertible, probably because one or more categories had too fewobservations. Treating wave 3 peer delinquency as a continuous variable, handling missing data (78 % foreach precursor) with FIML and running an ML analysis nevertheless produced results consistent with asignificant indirect effect for both delinquency (estimate = 0.003, 95 % BCBCI = 0.001–0.006) and reactivecriminal thinking (estimate = 0.001, 95 % BCBCI = 0.000*–0.003: *dividing the independent variable by 10revealed that this lower limit was > 0.000). In these analyses, the standardized coefficients for the a and b pathsof the delinquency indirect effect were 0.07 (p < 0.001) and 0.10 (p < 0.001), respectively, and the standardizedcoefficients for the a and b paths of the reactive criminal thinking indirect effect were 0.07 (p < .001) and 0.04(p < .05), respectively.

222 G. D. Walters

however, in evaluating the merits of this argument. First, single indirect effects arenearly always small [38]. Notions of full and partial mediation have been replaced bythe realization that nearly all mediation is partial and that single variables, at least insocial science research, normally account for only a small portion of the total effect [39,56]. Second, there was a 2-year gap between assessments in this study. There is a greatdeal that can happen in the course of a 2-year period to conceal and complicate variablerelationships, particularly when working with children and early adolescents where

Table 4 Results of a mediation path analysis of peer rejection as a predictor of peer delinquency: two-mediator model from study 2

Predictor b (95 % CI) β t p


Age −0.018 (−0.082, 0.043) −0.009 −0.58 .559

Race 0.060 (0.000, 0.125) 0.031 1.89 .058

Sex 0.019 (−0.043, 0.081) 0.010 0.61 .544

Delinquency-2 (outcome measure)

Peer rejection-1 0.112 (0.055, 0.180) 0.067 3.50 <.001

Age 0.176 (0.068, 0.279) 0.054 3.25 .001

Race 0.351 (0.242, 0.458) 0.107 6.32 <.001

Sex −0.409 (−0.523, −0.298) −0.125 −7.22 <.001

RCT-2 (outcome measure)

Peer rejection-1 0.202 (0.084, 0.328) 0.058 3.26 .001

Age 0.376 (0.156, 0.601) 0.055 3.30 <.001

Race −0.269 (−0.495, −0.046) −0.040 −2.36 .018

Sex −0.731 (−0.953, −0.502) 0.108 −6.36 <.001

Peer delinquency-3 (outcome measure)

Delinquency-2 0.070 (0.042, 0.097) 0.109 5.09 <.001

RCT-2 0.019 (0.003, 0.034) 0.062 2.44 .015

Peer rejection-1 0.148 (0.102, 0.191) 0.138 6.53 <.001

Peer delinquency-1 0.833 (0.718, 0.947) 0.286 14.03 <.001

Age 0.048 (−0.047, 0.146) 0.023 0.98 .329

Race 0.129 (0.026, 0.223) 0.061 2.62 .009

Sex 0.031 (−0.069, 0.130) 0.015 0.62 .534

Delinquency-2 with RCT-2 1.036 (0.842, 1.239) 0.192 10.24 <.001

N = 3774 (43 cases with missing data on x-variables were not included in the analysis)

Peer rejection-1 (outcome measure) regression equation with peer rejection at wave 1 as the predictedvariable, Delinquency-2 (outcome measure) regression equation with own delinquency at wave 2 as thepredicted variable, RCT-2 (outcome measure) regression equation with reactive criminal thinking at wave 2 asthe predicted variable, Peer delinquency-3 (outcome measure) regression equation with peer delinquentassociations at wave 3 as the predicted variable, Age chronological age in years, Race 1 (white), Race 2(non-white), Sex 1 (male), Sex 2 (female), Peer rejection-1 peer rejection at wave 1, Delinquency-2 owndelinquency at wave 2, RCT-2 reactive criminal thinking at wave 2, Peer delinquency-1 peer delinquentassociations at wave 1, Delinquency-2 with RCT-2 correlation between peer delinquency and reactive criminalthinking at wave 2, b (95 % CI) unstandardized coefficient and the lower and upper limits of the 95 %confidence interval for the unstandardized coefficient (in parentheses), β standardized coefficient, t asymptotict test, p significance level of the asymptotic t test


change is the norm. Third, the current study involved correlating variables fromdifferent domains given that the independent and dependent variables were based onmaternal ratings of child behavior and the mediators were based on child self-report.Correlations across domains are nearly always smaller than correlations within adomain [57], so it is somewhat misleading to compare the current study to the bulkof criminological research where correlations come from a single domain (normallyself-report).

Table 5 Total, direct, and indirect effects for pathways running from peer rejection to peer delinquency: twomediators from study 2

Pathways BCBCI

Estimate Lower Upper

Total effect 0.160 0.115 0.203

Direct effect 0.148 0.102 0.191

Total indirect effect 0.012 0.006 0.019

Specific indirect effects

Peer rejection → delinquency → peer delinquency 0.008 0.004 0.014

Peer rejection → RCT → peer delinquency 0.004 0.001 0.009


Peer rejection peer rejection at wave 1, Delinquency self-reported delinquency at wave 2, Peer delinquencypeer delinquent associations at wave 3, RCT reactive criminal thinking at wave 2, BCBCI bias-correctedbootstrapped 95 % confidence interval (b = 5000), Estimate unstandardized point estimate, Lower lowerboundary of the 95 % confidence interval, Upper upper boundary of the 95 % confidence interval

Wave 1 Wave 2 Wave 3

Peer Reject Peer Del.14**/.13**

Del

RCT

Depression

Fig. 3 Path analysis of the mediating effects of wave 2 delinquency, reactive criminal thinking, anddepression on the wave 1 peer rejection–wave 3 peer delinquency relationship for the two-mediator(bolded coefficients in front of slash) and three-mediator (non-bolded coefficients) models in study 2.Note. Standardized beta coefficients are reported; solid lines = pathways for the two- and three-mediatormodels; dashed lines = pathways for the third mediator (depression) of the three-mediator model; PeerReject = peer rejection at wave 1; Del = own delinquency at wave 2; RCT = reactive criminal thinking atwave 2; Depression = depression at wave 2; Peer Del = peer delinquent associations at wave 3; controlvariables are not shown. *p < 0.05, **p < 0.001

224 G. D. Walters

General Discussion

Results from the two studies reported in this paper offer preliminary support for severalpathways in the control model of criminal lifestyle development. Several earlier studiesprovisionally confirmed pathways in the front end, back end, and bottom portions ofthe model but the middle portion had yet to be tested. As a means of review, priorresearch on the front end of the model indicated that low self-control predicts reactivecriminal thinking which, in turn, predicts delinquency [10]. Regarding the back end ofthe model, there are now two studies [20, 58] showing that the peer influence effect(delinquent peer associations leading to participant delinquency) is mediated by pro-active criminal thinking. With respect to the lower portion of the model, there is nowpreliminary evidence supporting the psychological inertia theorem (delinquency lead-ing to reactive criminal thinking leading to delinquency [21]) and the peer selectioneffect (participant delinquency leading to reactive criminal thinking leading to delin-quent peer associations [20]). The current study corroborates most of the remainingpathways by documenting preliminary causal pathways from low self-control to peerrejection (study 1) and from peer rejection to delinquent peer associations, both directlyand via delinquency and reactive criminal thinking (study 2).

A theoretical implication of the current results is that conceptual integration ispossible despite skepticism by those who believe competition between theories is thebest way to promote science (e.g., [59]). Four different theoretical traditions (low self-control, general strain, differential association, and criminal thinking), each with itsown unique assumptions and methodologies, are represented in the control model ofcriminal lifestyle development (see Fig. 1) and each received preliminary support in thecurrent series of studies. The concerns of Wheeldon et al. [7] about integrating modelswith different underlying assumptions (e.g., free will versus determinism) and

Table 6 Total, direct, and indirect effects for pathways running from peer rejection to peer delinquency: threemediators from study 2

Pathways BCBCI

Estimate Lower Upper

Total effect 0.160 0.115 0.203

Direct effect 0.144 0.097 0.187

Total indirect effect 0.016 0.007 0.027

Specific indirect effects

Peer rejection → delinquency → peer delinquency 0.008 0.003 0.014

Peer rejection → RCT → peer delinquency 0.003 0.001 0.009

Peer rejection → depression → peer delinquency 0.005 −0.002 0.014


Peer rejection peer rejection at wave 1, Delinquency self-reported delinquency at wave 2, Peer delinquencypeer delinquent associations at wave 3, RCT reactive criminal thinking at wave 2, Depression depression atwave 2, BCBCI bias-corrected bootstrapped 95 % confidence interval (b = 5000), Estimate unstandardizedpoint estimate, Lower lower boundary of the 95 % confidence interval, Upper upper boundary of the 95 %confidence interval


methodologies (e.g., quantitative versus qualitative) notwithstanding, this initial attemptat theoretical integration achieved some degree of success. This may be because noassumptions need to be made about whether low self-control is innate or learned, agentic,or determined, or whether crime should be studied quantitatively or qualitatively, in orderfor the pathways to do their job. By extending the model backwards, however, we enter arealm where assumptions are vital. Is low self-control a function of biological factors liketemperament [10, 60] or environmental factors like parental support and control [4] arequestions rich in assumptions. These are also questions that are best answered empiricallywith qualitative and quantitative research. What the current results suggest is that one wayof achieving integration is by creating links between constructs using the methodologies ofmoderation and mediation, and while moderation is not currently represented in Fig. 1, itcould be added as research in this area progresses.

A practical implication of the current results is that they point to avenues by whichdelinquency and crime might be prevented. Trentacosta and Shaw [61] determined thatadaptive emotional self-regulation strategies like active distraction in childhood re-duced antisocial behavior in early adolescence by lessening peer rejection in latechildhood. Parenting training has also been found to be effective in reducing conductdisorder and delinquency in children and early adolescents with low self-control whomay be at risk for peer rejection [62, 63]. In a review of the prevention literature onaggression, van Lier et al. [64] surmised that aggression and violent behavior were bestprevented by developing programs that target peer relationships. One of the morepromising approaches they came across was one in which aggressive and delinquentchildren were integrated into groups of conventional peers. This, of course, will bemost effective during the early stages of delinquency. Targeting low self-control, peerrejection, and such mediators as early delinquency and reactive criminal thinkingshould therefore be more effective than trying to discourage delinquent peer associa-tions after they have already formed. Gottfredson and Hirschi [4] maintain that whileabsolute self-control improves with age, relative self-control does not change after age8 or 9. Research, however, indicates that both absolute and relative self-control changemoderately after middle childhood and that intervention and prevention programs canbe of assistance in facilitating the development of self-control [65, 66].

Before discussing the limitations of the research presented in this paper, I would liketo discuss several of the strengths. First, the variables in both studies were measuredacross domains rather than within a single domain. The exclusive use of self-report inmeasuring the variables in a study is a fairly common practice in criminology, but it is apractice that leads to mono-operational bias and artificially inflated correlations viashared method variance [67]. In study 1, peer rejection was measured with peer ratingsfrom a sociometric survey and low self-control was assessed with teacher reports; instudy 2, the independent (peer rejection) and dependent (peer delinquency) variableswere based on maternal ratings and the mediator variables were based on self-report. Asecond strength of both studies is that they controlled for continuity in the cross-laggedvariables by including the precursor to the predicted variable in both cross-laggedregressions while controlling for prior levels of the dependent variable in the two- andthree-mediator analyses for study 2. A further strength of study 2 is that it incorporateda control mediator into the three-mediator analysis designed to test the specificity of thegeneral strain effect. Prior research has shown that anger is significantly more likely tolead to a criminal adaptation than non-anger negative affect [23–25]. When a non-anger

226 G. D. Walters

negative affect measure of depression was added to the three-mediator model, itcorrelated with peer rejection but not with peer delinquency, it did not alter thesignificance of the two target mediators (delinquency and reactive criminal thinking),and it failed to achieve a significant indirect effect.

Turning our attention to limitations we can see that external validity was weak inboth studies. The sample for study 1 was a census of boys from seven schools locatedwithin a mile of the Cambridge study’s field research office. The sample for the secondstudy was comprised of boys and girls born to women from the nationally representa-tive NLSY-79 sample but was nothing more than a non-probability conveniencesample. Additional research is consequently required to test the generalizability ofthe results from both studies. A second limitation of both studies is the 2-year gapbetween each of the evaluation periods (waves) in studies 1 and 2. A great deal cantranspire in 2 years, particularly in samples of children and early adolescents wherechange can be both rapid and abrupt. Such a relatively long period of time betweenassessments could provide opportunities for non-study variables to infiltrate the designand influence the results in unanticipated ways. A third limitation is the absence ofmediators in study 1 and the large amount of missing data (>75 %) for the two-mediatorprecursors in study 2. Without precursor variables, it is impossible to rule out thealternate hypothesis that pre-existing differences in delinquency or reactive criminalthinking were the cause rather than the effect of peer rejection. Treating the three-leveldependent variable as a continuous measure and using FIML to compensate for thelarge amount of missing data in the mediator precursor variables did not appreciablyalter the results from those obtained in analyses conducted without mediator precursors(see footnote 1), although this may have been too much to ask even from FIML. Afourth limitation is that peer delinquency in study 2 was measured with a one-itemmaternal rating. Even though this item was used previously by Chapple [30] to assesspeer delinquency, a single item is subject to misinterpretation and idiosyncraticresponding. Research is also required to determine whether these results replicate whenpeer delinquency is measured with peer network nominations [68] rather than maternalratings.

On the subject of limitations, one could take issue with the construct validity of theself-report items used to assess reactive criminal thinking in study 2 given that thesesame items were used by Turner and Piquero [69] to measure low self-control. I wouldlike to explain, however, why I believe these items do a better job of assessing reactivecriminal thinking than they do of assessing low self-control. First, Gottfredson andHirschi [4] clearly state that low self-control should be assessed behaviorally, with ameasure like the BPI. Second, cognitive and perceptual factors like impression man-agement and self-serving bias are more likely to influence responding on a self-reportinventory than they are to shape responses to a behavioral rating scale [70, 71].Cognitive influence is present even when constructing a self-report measure of specificbehaviors (e.g., number of offenses committed in the past year), although the influenceis much stronger when an individual is asked to make subjective judgments about his orher experiences, opinions, and priorities. Self-serving bias and impression managementare actually part of the process of reactive criminal thinking. The fact that the sixreactive criminal thinking items used in study 2 correlated 0.43 with an establishedmeasure of reactive criminal thinking in a different sample of participants [48] lendsfurther credence to the construct validity of the RCT scale. Moreover, considering that


these items correlated only modestly with concurrent total BPI scores used in theTurner and Piquero [69] study (r=0.00–0.22, M= .15) and were three times less stablethan the BPI externalizing score in the Walters [10] investigation denotes that they aremeasuring something other than Gottfredson and Hirschi’s [4] behavioral self-controlconcept. Finally, results from a recent meta-analysis showed that self-report andbehavioral indicators of “low self-control” are measuring different constructs [72],perhaps because behavioral indicators are measuring low self-control and self-reportindicators are measuring reactive criminal thinking.

Results from studies conducted on peer rejection in both the Cambridge Study inDelinquent Development and NLSY-C indicate that constructs from general straintheory may play a role in the control model of criminal lifestyle development. Futureresearch could extend this model by replacing peer rejection with other putative causesof strain, such as parental conflict, school failure, and loss of freedom throughincarceration. The notion that only certain emotional concomitants of strain, anger inparticular, lead to criminal outcomes may well be another piece to the crime puzzle. Inthe current study, the direct effect of peer rejection on delinquent peer associations wasstill strong after accounting for the indirect effects of delinquency and reactive criminalthinking. This suggests that additional variables, like anger and hostile attributionbiases, may be involved in mediating the relationship between peer rejection and peerdelinquency, although a direct effect, or even a shared method variance effect [67], forthat matter, between peer rejection and delinquent peer associations cannot be ruled out.All of these possibilities are fertile soil for future research. In creating a path for futureresearch on the control model, one could argue that instead of evaluating the model inpiecemeal fashion, as was done in the current study, all of the pathways should bestudied simultaneously. Although well-intended, this advice is impractical given therelatively low precision of current key measures and the even lower power of currentstatistical methods to test more than three or four pathways at a time (keeping in mindthat overall model fit is not a legitimate means of evaluating indirect effects). There are15 paths and 7 sequences currently represented in the control model. Before we caninstantaneously put all of the pieces of this puzzle together, the precision of ourmeasures and the power of our methodologies would need to be significantly enhanced.

References

1. Bernard, T. J., & Snipes, J. B. (1996). Theoretical integration in criminology. Crime and Justice: A Reviewof Research, 20, 301–348.

2. Ericson, R., & Carriere, K. (1994). The fragmentation of criminology. In D. Nelken (Ed.), The futures ofcriminology (pp. 89–109). London: Sage.

3. Sutherland, E. H. (1947). Principles of criminology (4th ed.). Philadelphia: J.B. Lippincott.4. Gottfredson, M. R., & Hirschi, T. (1990). A general theory of crime. Stanford: Stanford University Press.5. Agnew, R. (2011). Toward a unified criminology: integrating assumptions about crime, people, and

society. New York: New York University Press.6. Thornberry, T. P. (1987). Toward an interactional theory of delinquency. Criminology, 25, 863–892.7. Wheeldon, J., Heidt, J., & Dooley, B. (2014). The trouble(s) with unification: debating assumptions,

methods, and expertise in criminology. Journal of Theoretical and Philosophical Criminology, 6, 111–128.

8. Liska, A. E., Krohn, M. D., & Messner, S. F. (1989). Strategies and requisites for theoretical integration inthe study of crime and deviance. In S. F. Messner, M. D. Krohn, & A. E. Liska (Eds.), Theoretical

228 G. D. Walters

integration in the study of deviance and crime: problems and prospects (pp. 1–18). Albany: StateUniversity of New York Press.

9. Walters, G. D. (2012). Crime in a psychological context: from career criminals to criminal careers.Thousand Oaks: Sage.

10. Walters, G. D. (2015). Early childhood temperament, maternal monitoring, reactive criminal thinking, andthe origin(s) of low self-control. Journal of Criminal Justice, 43, 369–376.

11. Walters, G. D., & Kiehl, K. A. (2015). Limbic correlates of fearlessness and disinhibition in incarceratedyouth: exploring the brain-behavior relationship with the Hare Psychopathy Checklist: Youth Version.Psychiatry Research, 230, 205–210.

12. Meldrum, R. C., Young, J. T. N., & Lehmann, P. S. (2015). Parental low self-control, parental socializa-tion, young adult low self-control, and offending. Criminal Justice and Behavior, 42, 1183–1199.

13. Hay, C. (2001). Parenting, low self-control, and delinquency: a test of self-control theory. Criminology, 39,707–736.

14. Pratt, T. C., & Cullen, F. T. (2000). The empirical status of Gottfredson and Hirschi’s general theory ofcrime: a meta-analysis. Criminology, 38, 931–964.

15. Mathias, C. W., Marsh-Richard, D. M., & Dougherty, D. M. (2008). Behavioral measures of impulsivityand the law. Behavioral Sciences and the Law, 26, 691–707.

16. Kroner, D. G., & Morgan, R. D. (2014). An overview of strategies for the assessment and treatment ofcriminal thinking. In R. C. Tafrate & D. Mitchell (Eds.), Forensic CBT: a handbook for clinical practice(pp. 87–101). Chichester, England: Wiley.

17. Pratt, T. C. (2016). A self-control/life-course theory of criminal behavior. European Journal ofCriminology, 13, 129–146.

18. Walters, G. D. (in press). Reactive criminal thinking as a consequence of low self-control and prioroffending. Deviant Behavior.

19. Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs: Prentice Hall.20. Walters, G. D. (in press). Friends, cognition, and delinquency: proactive and reactive criminal thinking as

mediators of the peer influence and peer selection effects among male delinquents. Justice Quarterly.21. Walters, G. D. (in press). Crime continuity and psychological inertia: testing the cognitive mediation and

additive postulates with male adjudicated delinquents. Journal of Quantitative Criminology,22. Agnew, R. (1992). Foundation for a general strain theory. Criminology, 30, 47–87.23. Broidy, L. (2001). A test of general strain theory. Criminology, 39, 9–33.24. Jang, S. J. (2007). Gender differences in strain, negative emotions, and coping behaviors: a general strain

theory approach. Justice Quarterly, 24, 523–549.25. Scheuerman, H. L. (2013). The relationship between injustice and crime: a general strain theory approach.

Journal of Criminal Justice, 41, 375–385.26. Agnew, R. (2013). When criminal coping is likely: an extension of general strain theory. Deviant

Behavior, 34, 653–670.27. Turanovic, J. J., & Pratt, T. C. (2013). The consequences of maladaptive coping: integrating general strain

and self-control theories to specify a causal pathway between victimization and offending. Journal ofQuantitative Criminology, 29, 321–345.

28. Higgins, G. E., Piquero, N. L., & Piquero, A. R. (2011). General strain theory, peer rejection, anddelinquency/crime. Youth and Society, 43, 1272–1297.

29. Bolger, K. E., & Patterson, C. J. (2001). Developmental pathways from child maltreatment to peerrejection. Child Development, 72, 549–568.

30. Chapple, C. L. (2005). Self-control, peer relations, and delinquency. Justice Quarterly, 22, 89–106.31. Olson, S. L., & Hoza, B. (1993). Preschool developmental antecedents of conduct problems in children

beginning school. Journal of Clinical Child Psychology, 22, 60–67.32. Coie, J. D., Lochman, J. E., Terry, R., & Hyman, C. (1992). Predicting early adolescent disorder from

childhood aggression and peer rejection. Journal of Consulting and Clinical Psychology, 60, 783–792.33. Fergusson, D. M., & Horwood, L. J. (1999). Prospective childhood predictors of deviant peer affiliations

in adolescence. Journal of Child Psychology and Psychiatry, 40, 581–592.34. Guerra, V. S., Asher, S. R., & DeRosier, M. E. (2004). Effect of children’s perceived rejection on physical

aggression. Journal of Abnormal Child Psychology, 32, 551–563.35. Miller-Johnson, S., Coie, J. D., Maumary-Gremaud, A., Lochman, J., & Terry, R. (1999). Relationship

between childhood peer rejection and aggression and adolescent delinquency severity and type amongAfrican American youth. Journal of Emotional and Behavioral Disorders, 7, 137–146.

36. Lansford, J. E., Dodge, K. A., Fontaine, R. G., Bates, J. E., & Pettit, G. S. (2014). Peer rejection, affiliationwith deviant peers, delinquency, and risky sexual behavior. Journal of Youth and Adolescence, 43, 1742–1751.


37. Vitaro, F., Pedersen, S., & Brendgen, M. (2007). Children’s disruptiveness, peer rejection, friends’deviancy, and delinquent behaviors: a process-oriented approach. Development and Psychopathology,19, 433–453.

38. Rucker, D. D., Preacher, K. J., Tormala, Z. L., & Petty, R. E. (2011). Mediation analysis in socialpsychology: current practices and new recommendations. Social & Personality Psychology Compass, 5,359–371.

39. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: aregression-based approach. New York: Guilford.

40. West, D. J., & Farrington, D. P. (1973). Who becomes delinquent? London: Heinemann.41. Muthén, B., & Muthén, L. (1998–2007).Mplus user’s guide (5th ed.). Los Angeles: Muthén and Muthén.42. Allison, P. D. (2012) Handling missing data by maximum likelihood. SAS Global Forum 2012 (Paper

312–2012). Cary, NC: SAS Institute43. Peyre, H., Leplége, A., & Coste, J. (2011). Missing data methods for dealing with missing items in quality

of life questionnaires: a comparison by simulation of personal mean score, full information maximumlikelihood, multiple imputation, and hot deck techniques applied to the SF-36 in the French 2003decennial health survey. Quality of Life Research, 20, 287–300.

44. Raghunathan, T. E., Rosenthal, R., & Rubin, D. B. (1996). Comparing correlated but nonoverlappingcorrelations. Psychological Methods, 1, 178–183.

45. Agnew, R. (2001). Building on the foundation of general strain theory: specifying the types of strain mostlikely to lead to crime and delinquency. Journal of Research in Crime and Delinquency, 38, 319–361.

46. Center for Human Resource Research. (2009). NLSY79 user’s guide, CHRR NLS User Services, TheOhio State University, Columbus, OH

47. Peterson, J. L., & Zill, N. (1986). Marital disruption, parent–child relationships, and behavioral problemsin children. Journal of Marriage and Family, 48, 295–307.

48. Walters, G. D., & Yurvati, E. (2015). Testing the construct validity of the PICTS proactive and reactivescores against six putative measures of proactive and reactive criminal thinking.Manuscript submitted forpublication.

49. Walters, G. D. (1995). The Psychological Inventory of Criminal Thinking Styles: part I. Reliability andpreliminary validity. Criminal Justice and Behavior, 22, 307–325.

50. Walters, G. D., Hagman, B. T., & Cohn, A. M. (2011). Towards a hierarchical model of criminal thinking:evidence from item response theory and confirmatory factor analysis. Psychological Assessment, 23, 925–936.

51. Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data: questions andtips in the use of structural equation modeling. Journal of Abnormal Psychology, 112, 558–577.

52. MacKinnon, D. P., Kisbu-Sakarya, Y., & Gottschall, A. C. (2013). Developments in mediation analysis. InT. D. Little (Ed.), The Oxford handbook of quantitative methods (Vol. 2, pp. 338–360). New York: OxfordUniversity Press.

53. Pituch, K. A., & Stapleton, L. M. (2008). The performance of methods to test upper-level mediation in thepresence of nonnormal data. Multivariate Behavioral Research, 43, 237–267.

54. Kenny, D. A. (2013). Mediation: sensitivity analysis [webinar]. Retrieved from http://davidakenny.net/webinars/Mediation/Sensitivity/Sensitivity.html.

55. Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparingindirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891.

56. Preacher, K. J. (2015). Advances in mediation analysis: a survey and synthesis of new developments.Annual Review of Psychology, 66, 825–852.

57. Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., & Reed, G. M. (2001). Psychologicaltesting and psychological assessment: a review of evidence and issues. American Psychologist, 56, 128–165.

58. Walters, G. D. (2015). Proactive criminal thinking and the transmission of differential association: a cross-lagged multi-wave path analysis. Criminal Justice and Behavior, 42, 1128–1144.

59. Hirschi, T. (1989). Exploring alternatives to integrated theory. In S. F. Messner, M. D. Krohn, & A. E.Liska (Eds.), Theoretical integration in the study of deviance and crime: problems and prospects (pp. 37–49). Albany: State University of New York Press.

60. DeLisi, M., & Vaughn, M. G. (2014). Foundation for a temperament-based theory of antisocial behaviorand criminal justice system involvement. Journal of Criminal Justice, 42, 10–25.

61. Trentacosta, C. J., & Shaw, D. S. (2009). Emotional self-regulation, peer rejection, and antisocialbehavior: developmental associations from early childhood to early adolescence. Journal of AppliedDevelopmental Psychology, 30, 356–365.

230 G. D. Walters

http://davidakenny.net/webinars/Mediation/Sensitivity/Sensitivity.html

http://davidakenny.net/webinars/Mediation/Sensitivity/Sensitivity.html

62. Dretzke, J., Davenport, C., Frew, E., Barlow, J., Stewart-Brown, S., & Hyde, C. (2009). The clinicaleffectiveness of different parenting programmes for children with conduct problems: a systematic reviewof randomized controlled trials. Child and Adolescent Psychiatry and Mental Health, 3, 7.

63. Lundahl, B., Risser, H. J., & Lovejoy, M. C. (2006). A meta-analysis of parent training: moderators andfollow-up effects. Clinical Psychology Review, 26, 86–104.

64. van Lier, P., Vitaro, F., & Eisner, M. (2007). Preventing aggressive and violent behavior: using preventionprograms to study the role of peer dynamics in maladjustment problems. European Journal on CriminalPolicy and Research, 13, 277–296.

65. Hay, C., Meldrum, R., Forrest, W., & Ciaravolo, E. (2010). Stability and change in risk seeking:investigating the effects of an intervention program. Youth Violence and Juvenile Justice, 8, 91–106.

66. Koning, I. M., Verdurmen, J. E. E., Engels, R. C. M. E., van den Eijnden, R. J. J. M., & Vollebergh, W. A.M. (2012). Differential impact of a Dutch alcohol prevention program targeting adolescents and parentsseparately and simultaneously: low self-control and lenient parenting at baseline predict effectiveness.Prevention Science, 13, 278–287.

67. Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-experimental designs forgeneralized causal inference. Boston: Mifflin.

68. Haynie, D. L., & Osgood, D. W. (2005). Reconsidering peers and delinquency: how do peers matter?Social Forces, 84, 1109–1130.

69. Turner, M. G., & Piquero, A. R. (2002). The stability of self-control. Journal of Criminal Justice, 30, 457–471.

70. Henry, D. B. (2006). Associations between peer nominations, teacher ratings, self-reports, and observa-tions of malicious and disruptive behavior. Assessment, 13, 241–252.

71. Peets, K., & Kikas, E. (2006). Aggressive strategies and victimization during adolescence: grade andgender differences, and cross-informant agreement. Aggressive Behavior, 32, 68–79.

72. Walters, G. D. (in press). Are behavioral measures of self-control and the Grasmick self-control scalemeasuring the same construct? A meta-analysis. American Journal of Criminal Justice.


Low Self-Control, Peer Rejection, Reactive Criminal Thinking, and … · 2017-08-26 · Low...

Documents

Transcript of Low Self-Control, Peer Rejection, Reactive Criminal Thinking, and … · 2017-08-26 · Low...